Salut Krako,
Merci pour ton message. J'en pense qu'effectivement les tags que tu propose sont plus pertinents que ceux que j'ai utilisé.
Pour t'expliquer la configuration de ce terrain, il s'agit de fond de cuvette oasienne, c'est une terre limoneuse/argileuse très chargé en diatomée. Le fond se transforme en mare bordée de boue durant la saison des pluies et durant quelques petits mois après... puis s'assèche complètement sauf dans quelque rare cas ou la cuvette est suffisamment profonde, et la nappe affleure en permanence ( et là l'étang est bien bordé de boue en permance).
si tu peux et le faire ce changement de tag, n'hésite pas, et merci de m'avoir contacté avant!

hi Mateusz. Thank for the tip, i will ad this tag next time i have to map spead residential areas. Map it more accurately is difficult : where is the limit of the reidential use and the agricultural use ? the "perfect" solution will be to map only the building. therfore this solution have for consequence that the populated area are visible only at the last zoom level, wath is quite useless for the map users. And this solution il also exessively time consuming. For large areas of africa, where the map is quite ''virgin'', the maper's time is a limited resource...

You are true. It would be more appropiate following the osm way of dooing. But as i say to datakoop, the nodle will note apear on the Map and common user '' osm consummer'' will not see the information that is an inhabited area...

Waouh! Many thanks. You juste withdrow a lot of houres of '' lassy'' work...
If you dont see the interest of this, in tout mind, what is the interest of a blank map?
For the local civil socity, that is very important to see where are the most populated area on the Map. And no of them Can affords spending ressources and time on dedicated shared mapping . A mater of live...
Yes, a few part of this work is probably wrong ( dis you check with the whole satimagery before ?) . You cant withdrow all juste because there is feew mistakes. Unless you have to delete the whole map.
What is the interest of a particiativ mapping if you dont care of the motivation of the mapper?
Deleting this informations is largely lassyer than trying improving it.

Hi Mr Atlas,
Indeed, actively practicing the roads of Senegal I found that OSM maps provided little discriminating indication of the quality of main and secondary roads, largely mapped in "unclassified", which does not allow an interesting use of OSM maps for travel. Many important roads and laterite tracks in Senegal are new and need updating.
The definition of road tags in the wiki is fine for Europe/USA, but for Africa to follow them strictly means to lose a lot of information when you are on the side of the data user. The surface=asphalt tag does not imply a difference in visibility on the map. Roads are rare compared to tracks that connect important cities, important not necessarily in size but in terms of geography of the territory.
Concerning the main axes in the cities, I plan to do it after having mapped the whole rural territory. (I sometimes overflowed a little to make the junction between main axes, it will be corrected at that time).
For the particular case of the "pipline", in absence of source, and by noticing connected laterite backs, I preferred to modify the tag in route, in coherence with the visual of the other similar roads. ( I imagine that the "pipeline" was tagged like this on a photo sat of road works)

Translated with www.DeepL.com/Translator (free version)

The circular blobs is the more effective way of representation that I found to those area. using a node show the same issue; where to place it ? for every house? in the barycenter of 10 houses? 100 houses? And the nodes are note visible on the map.
If people are interested to refine the map, it is easy for them to delet the shape ( it is not linked to other with multi-poligone relation). instead, the blobs could show them where they have to work without downloading large area of landscape.
by the way, those remote area do not show an attract for mappers and the map remain virgin for a long time...
"It is even worse than a blank map'', As a field pratictioner, I insure you that it is largely more useful to have a good aproximation of the reality than a blank map.
Note that i never make fake, but proxi. That make a difference.
finally, OSM Tags are designed for an occident context. the Africa rurality is far from this context. Boundaries are moving and unclear. African OSM mapping community is few. the task is huge. Allow us to not dedicate too much time to produce sharper maps: sharper they are, faster theye out-date.

hi Mateusz. Please, could you exlain why do you think the tag I used is not appropriate for the low density village i maped? if it'is the circular form of thoses area that make you inconfortable, could you tell me how to map the villages on this area ( what is the formal milit betwwen the residential, agricultural, Natural area,...? )

The area I map as ‘Landuse=residential ‘ is based on evidence of inhabitation : houses, farmyard, livestock Shed, fences,… but with no clear border. Even my work seem to be rough, you will have the same issue at lower scale.
A heavy work on tree mapping or house to house mapping in that kind of context, will not be more “true” with the next satellite imagery available.
You can make the same comment for every landuse mapped on African countryside.
OSM is a great tool, well adapted for western country context. Using it on this kind of place need to manage with inconstancy and blurred border.
What is better: a true proxy or a data lack (and lack is à kind of information, could be interpreted as there is no people/nothing here… what is largely wronged). The perfect circle y use for this proxy is the same level of true than official mapping of large scale map.
For evidence of usefulness, I currently work with Red Cross, on flood preparness on the area.
(https://unocha.us4.list-manage.com/track/click?u=547a787708d32a96a42c77746&id=01c5dee01f&e=be2e3f9de8)
Because of lack of easy access Data, available on mobile support, our team focused on a wrong area for 2 years, working near Bongor. Thanks to my work ( and the other few OSM contributor), they are now focusing in the most vulnerable area to flood, and more exposed people.
Lazzy is a tought judgment for what I doo, covering large area with essential information take time.
For myself I do not emit judgment for people “wasting” time on useless overdetailed data while the map is steel virgin on the whole huge area around.
I spend time to response in detail to you 2 sentences in a foreign language for me. If you are steel skeptical, please take time to understand the specific context I map before charging my work as useless. Thanks and have a nice day

Dear DataKop,
You question seem to be rhetorical, ;-) = If I spend time for this, that mean that it’s worth the time spend, isn’t? …
To respond to the question “ Why maping ‘landuse=residential’ like this ? “
I think the two last words in you first question are the more important
- “Later” : on a large part of the map where I contribute, later is close to never…
During this time, the map shows NO indication about populated area.
When somebody will refine it later, the population growth will probably void a current “refined” mapping.
And if somebody has time to spend on it, my “fastmaping” will help to show him where is the populated Area
The use of tag ‘place: village’ or ‘ place: hamlet’ without name is not visible on the map so not very useful.

- “refined” : on this area, there is no clear border to land ownership, or clear border between the house area, the backyard crop garden, and the fields. So using a circle is not a bad way to show the information.

I hope that have respond to you implicit question, and begin to convince you to the interest of this kind of premaping / fastmaping.
PS: for my sector (humanitarian), OSM is the main reference for cartography, and the information I map is useful for our team on the field, to priories the main populated area impacted by a disaster, aiming objective choice.

Bonsoir Rafael,
J'ai vu le fil de discutions, il y a des points de vus différents...
- une proposition de demander ce qu'en pense les concernés : la communauté OSM de RCA : sont-ils actifs ? Un bon proxy serait de demander à la communauté Cameroun, ou de voir leurs choix directement sur la carte ;-)
- A mon sens je trouve qu’il manque le point de vue de l’utilisateur de la carte. Si on contribue autant sur OSM, c’est pour in fine, en faire des cartes utiles pour tous non ?
On a des avis de cartographe, mais peut-on se projeter dans la tête d’un usager ? Comment, sans connaitre le terrain, peut-il sur la carte trouver des infos qui lui seront utile en un coup d’œil? Qui ne donne pas de fausses indications ? Ou qui ne regroupe pas sous la même apparence des réalités bien différentes

(Je n’interviens pas sur les échanges dans le wiki, car pas très à l’aise pour argumenter en anglais et ne maitrise pas toute les subtilités de OSM… je ne suis qu’un modeste contributeur sans légitimité autre que ma connaissance de l'endroit en question…)
Pierre

Parfois j'ai aussi utiliser Rank=0 pour faire ressortir certaines lieux non administratif mais très importante pour la navigation, comme certains lieu dans le desert, qui marque un point d'eau, une étape importante sur la piste, des lieux clés... qui ont leur utilité sur les cartes a des niveau de zoom haut. un peu comme le "puits de la dernière chance" au milieu du désert...

merci pour cette explication claire, je ne l'ai trouvé nulle part si explicite sur le wiki !
L'objectif initial de l'utilisation de ce Tag est de permetre de voir les villes à un niveau de zoom "haut", afin de faciliter la navigation sur les cartes OSM et avec les apli type Maps-Me. Les villages et town apparaissent sinon a un niveau de zoom qui ne permet pas appréhender une zone suffisement grande pour s'orienter ( surtout sur de petit écrans de smartphone sur le terrain...).
Avec le tag Capital=N, on arrive à ce même rendu graphique? si oui, parfait, je reprendrai l'ensemble du travail RCA+Tchad+Cameroun pour cleaner ça et ajouter ce Tag Capital
sinon, il faudrait l'ajouter (du coup je le ferais aussi) mais conserver le Rank=0 serait précieux pour les utilisateur.

C'est plus pour l'importance économique de l'axe Garoua-Boulaye --> Bangui que je l'ai tagé en Trunk,
comme pour le Cameroun, où le même tag est utilisé pour Douala-->Youandé-->Maroua (alors que ce ne sont pas des 2x2 voies non plus)
Ensuite c'est parce qu'il a cet importance économique qu'il a été goudronné... ça se recoupe.
La communauté OSM en RCA, je connais pas, mais Willy Zob de OSM Cameroun y va te temps en temps..
(vu la connectivité à Bangui, il faut du courage pour être un contributeur OSM en RCA !)

Bonjour Edvac,
Effectivement la clée "trunk" n'est pas tout a fait adapté si l'on suit wiki.
Cependant, si j'ai pris la peine de modifier les axes Bangui-Douala - Bangui Sibut et Banqui- Mbaili, c'est par soucis de visibilité de ces axes importants et goudronés ( les seuls de RCA) afin de souligner leur importance par rapports aux autres axes primaires, non goudronné. Aussi, ce même choix a été fait par d'autre contributeur dans les pays voisins pour lesquelles je contribue (Cameroun et Tchad) pour marquer les axes les plus circulant. OSM est un excellent outils, mais encore mal adapté aux réalités africaines sur beaucoup d'aspect... si vous avez des moyens alternatif pour faire ressortir ce besoin de visibilité de ces axes je suis preneur...

Hi Flynn,
I have try to fix this, could you tel me if it's ok now?